Do NIR spectra collected from laboratory-reared mosquitoes differ from those collected from wild mosquitoes?

Authors: MILALI, MASABHO - IFAKARA HEALTH INSTITUTE; SIKULU-LORD, MAGGY - UNIVERSITY OF QUEENSLAND; KIWARE, SAMSON - IFAKARA HEALTH INSTITUTE; DOWELL, FLOYD; POVINELLI, RICHARD - MARQUETTE UNIVERSITY; CORLISS, GEORGE - MARQUETTE UNIVERSITY

Submitted to: PLOS ONE
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 5/16/2018
Publication Date: 5/31/2018
Citation: Milali, M.P., Sikulu-Lord, M.T., Kiware, S.S., Dowell, F.E., Povinelli, R.J., Corliss, G.F. 2018. Do NIR spectra collected from laboratory-reared mosquitoes differ from those collected from wild mosquitoes? PLoS One. 13(5):e0198245. https://doi.org/10.1371/journal.pone.0198245.
DOI: https://doi.org/10.1371/journal.pone.0198245

Interpretive Summary: Several studies report that near infrared spectroscopy (NIRS) can estimate the age of lab-reared mosquitoes into either less than or greater than seven days old with an accuracy exceeding 80%. The ability of NIRS to estimate the age of laboratory and semi-field raised mosquitoes is a prerequisite for accurately predicting the age of wild mosquito samples. It is challenging to develop a NIRS model that can estimate the age of wild mosquitoes because it is difficult and expensive to get wild mosquitoes of a known age in days with which to train and validate the model. A model trained on spectra from laboratory-reared mosquitoes where age in days is known can be applied to estimate the age of wild mosquitoes, but this would be appropriate if spectra collected from laboratory-reared mosquitoes are similar to those collected from wild mosquitoes, as this study demonstrates. We performed cluster analysis on a mixture of spectra collected from lab-reared and wild Anopheles arabiensis to determine if there is any significant difference between spectra from the two groups. While controlling the age of mosquitoes, we found two clusters with no significant difference in distribution of spectra collected from lab-reared and wild mosquitoes. We repeated the analysis using hierarchical clustering, and similarly, no significant difference was observed. Therefore, we find no difference between spectra collected from laboratory-reared and wild mosquitoes of the same age and species, suggesting that applying the model trained on laboratory-reared mosquitoes might be appropriate.

Technical Abstract: Do near infrared spectra from lab-reared mosquitoes differ from spectra from wild mosquitoes? Near infrared spectroscopy (NIRS) can classify the age of lab-reared mosquitoes as younger or older than seven days with accuracy greater than 80%. Hence, it has been proposed in several studies as a complementary method to Detinova ovary dissection. For NIRS to be able to estimate age of wild mosquitoes, a sample of wild mosquitoes with known age in days is required to train and test the model. Getting actual age in days of wild mosquitoes is almost impractical as it is very difficult, tedious, time inefficient, and expensive. Training a model using labels from Detinova dissection results in a model with poor accuracy. Alternatively, a model trained on spectra from laboratory-reared mosquitoes where age in days is known can be applied to estimate the age of wild mosquitoes, but this would be appropriate if spectra collected from laboratory-reared mosquitoes are similar to those collected from wild mosquitoes, as this study demonstrates. We performed k-means (k = 2) cluster analysis on a mixture of spectra collected from lab-reared and wild An. arabiensis to determine if there is any significant difference. While controlling the age of mosquitoes, we found two clusters with no significant difference in distribution of spectra collected from lab-reared and wild mosquitoes (P = 0.245). We repeated the analysis using hierarchical clustering, and similarly, no significant difference was observed (P = 0.129). Therefore, we find no difference between spectra collected from laboratory-reared and wild mosquitoes of the same age and species, suggesting that applying the model trained on laboratory-reared mosquitoes might be appropriate.